currency crises in reverse: do large real exchange rate ... · 2.1. definition of large real...
Post on 21-Jul-2020
2 Views
Preview:
TRANSCRIPT
Munich Personal RePEc Archive
Currency Crises in Reverse: Do Large
Real Exchange Rate Appreciations
Matter for Growth?
Bussière, Matthieu and Lopez, Claude and Tille, Cédric
Banque de France, Banque de France, Graduate Institute for
International and Development Studies
January 2013
Online at https://mpra.ub.uni-muenchen.de/44096/
MPRA Paper No. 44096, posted 31 Jan 2013 17:25 UTC
1
Currency Crises in Reverse:
Do Large Real Exchange Rate Appreciations Matter for Growth?∗
Matthieu Bussière♠ , Claude Lopez♦, and Cédric Tille♣
Abstract
While currency crises have been extensively studied, the opposite phenomenon,
large appreciations, has been far less researched. We fill this gap by providing an
empirical exploration of historical episodes of large real exchange rate
appreciations, using a sample of 28 advanced and 25 emerging market economies,
with annual data going back to 1970. We focus on the impact of large
appreciations on output growth. Our first finding is that countries experiencing
large real exchange rate appreciations display distinct patterns: large appreciations
significantly lower export growth and boost import growth on impact. Strikingly,
however, output growth is higher, on average, despite the adverse impact on
exports. Our second finding is that these aggregate numbers hide substantial
heterogeneity, which we link to the nature of the shocks that cause the
appreciation. In particular, appreciations associated with so-called “capital flow
bonanzas” have a marked downward effect on growth. This pattern is consistent
with the insights from a simple model that contrasts the impact of productivity
shocks with that of capital inflows shocks. Higher productivity in the traded sector
leads to a boom in traded output and a current account surplus, while higher
foreign lending leads to a boom in non-traded output and an external deficit as
traded output falls and consumption increases.
Keywords: exchange rate, currency crises, endaka, international trade,
international capital flows, lending booms, small open economy macroeconomics.
JEL classification: F10, F30, F41.
∗ The views presented in this paper are those of the authors and do not necessarily reflect those of the Banque de
France or the Eurosystem. We are grateful to Philippe Bacchetta, Christiane Baumeister, Bruno Cabrillac, Menzie
Chinn, James Hamilton, Lutz Kilian, Céline Mistretta and Robert Vigfusson for helpful comments and suggestions.
We also would like to thank Catherine Martin for outstanding research assistance. ♠ Banque de France, 31 rue Croix des Petits Champs, 75001 Paris; email: matthieu.bussiere@banque-france.fr.
♦ Banque de France, 31 rue Croix des Petits Champs, 75001 Paris; email: claude.lopez@banque-france.fr. ♣ Graduate Institute for International and Development Studies P.O. Box 136, 1211 Genève 21, Switzerland, email:
cedric.tille@graduateinstitute.
2
1. Introduction
Currency crises (large real exchange rate depreciations) have figured very prominently on the
research agenda of international macroeconomics for several decades. Their adverse effects have
been broadly documented, leading academics and policy makers to develop tools to detect when
the economy is at risk of a crisis and policies to limit their impact, such as limiting exchange rate
movements – a pattern known as “fear of floating” (Calvo and Reinhart, 2002).1
In contrast to the impact of large exchange rate depreciations, relatively little is known about the
effect of large appreciations, even though concerns about such appreciations are clearly
pervasive in the policy debate. Sharp appreciations can lead to a loss in competitiveness and
therefore may reduce growth through net trade. This fear has been repeatedly expressed by a
broad range of prominent policy makers, with selected quotes given in Appendix I.2. While the
recent concern about “currency wars” – where the expansionary monetary policy in advanced
economies leads to capital flows to emerging markets and an appreciation of their currencies –
has been widely noted, the concern is not limited to emerging markets. Japanese policy makers
for instance have repeatedly expressed their worries, having experienced several episodes of
large real exchange rate appreciations, in the early 1970s in the wake of the Smithsonian
Agreement, from 1985 to 1995 after the Plaza Accord, and in the current crisis.3 The concern is
however not universally shared. Paul Krugman (1994) has dubbed it “a dangerous obsession”
and argues that “concerns about competitiveness are, as an empirical matter, almost completely
unfounded”.
1 The adverse effects on growth are documented by, among others, Cerra and Saxena (2008), and Bussière, Saxena
and Tovar (2012). Examples of efforts to develop statistical tools aiming to detect currency crises in advance include
so-called early warning models (Kaminsky, Lizondo and Reinhart, 1998, Goldstein and Reinhart, 2000) and logit
models (Frankel and Rose, 1996, Eichengreen, Rose and Wyplosz, 1995, Berg and Pattillo, 1999, Bussière and
Fratzscher, 2006). 2 Appendix I reveals that viewpoints are not identical. We reproduce here two quotes that illustrate these different
views. On the one hand, Mishkin (2007) clearly expresses concern about the competitiveness channel (“An
appreciation of the dollar, in turn, restrains exports (because the price of U.S. goods rises when measured in foreign
currencies) and stimulates imports (because imports become cheaper in dollar terms). The resulting decrease in net
exports implies a reduction in aggregate demand”). On the other hand, Noyer (2007) brings a more balanced view:
“It is clear that the price-competitiveness of French industries has deteriorated significantly in recent years. Has the
euro’s appreciation played a role in this? On the one hand, it undoubtedly penalises export sectors whose
competitors are located in other monetary areas. But, on the other hand, it benefits those sectors which are large
consumers of imported commodities. At this stage, the overall effect on France’s growth and external balance is not
clearly apparent”. 3 There is even a special word in Japanese to refer to a period of strong appreciation (“endaka”).
3
This active debate raises the question of the observed impact of large appreciations: is it so clear
that growth suffers in countries whose currency appreciates, relative to other countries? To
investigate this question, we assemble a large dataset of 52 countries (28 advanced economies
and 25 emerging markets), with annual data from 1970 to 2011. We identify large appreciation
episodes by relying on the approach of Kappler et al. (2011). As we focus on large appreciations,
we do not consider that a nominal appreciation, in one given year, qualifies, but instead restrict
ourselves to sustained and large real exchange rate movements. We first present key stylized
facts on the effect of large appreciations using descriptive statistics and event-case analysis.
Second, we show how the empirical pattern is consistent with the results from a simple
theoretical model.
Our main empirical results are as follows. First, large appreciations are neither uncommon nor
limited to emerging economies. We identify 30 episodes among which about two-thirds took
place in advanced economies.4 Second, large appreciations are on average associated with lower
exports and higher imports compared to other time periods. This result would comfort policy
makers who worry about the effects of large appreciations on net trade. Our third finding is that
this ceteris paribus effect is however compensated by other factors, with the consequence that
output growth is on average higher during large appreciation episodes despite the adverse effect
on net exports. This result therefore puts fears about the growth impact of large appreciations in
perspective. Fourth, we document substantial heterogeneity across episodes. In particular, we
find that appreciations associated with large net capital inflows (capital flow bonanzas) and rapid
growth in domestic credit (“lending booms”) are characterized by lower output growth.
The empirical pattern can be accounted for by a simple open economy model with differentiated
traded and non-traded sectors. The Home country’s real exchange rate appreciates when
productivity increases in the Home traded sector or when the propensity to save in the rest of the
world increases, a proxy for capital flow bonanzas. A given appreciation can be associated with
very different growth patterns depending on the nature of the shock. Higher productivity in the
Home traded sector leads to higher output in the traded sector and overall, and a current account
surplus if the shock is temporary. Higher savings in the Foreign country by contrast lead to a
4 By contrast there are 95 instances of a nominal appreciation in a given year.
4
smaller increase in overall output, a large shift of output composition towards the non-traded
sector, and a current account deficit.
While the literature on large appreciation is thinner than the one on currency crises, we are by no
means the first paper to consider the issue. Kappler et al. (2011) were the first to formally define
a large exchange rate appreciation and to look at the effects of such episodes on the current
account balance and on real output. Compared to Kappler et al. (2011), we are both more
restrictive (to the extent that we do not look at the current account balance) and more
encompassing, as we consider cross-country heterogeneity in addition to average responses. To
understand why some countries faced with a large appreciation manage to grow at a robust pace,
while others seem to be particularly affected, we pay particular attention to the cases of large
appreciations associated with “capital flow bonanzas” (Reinhart and Reinhart (2008)) or with
“lending booms” (Gourinchas et al. (2001)), a focus that to our knowledge has not been taken
previously.
Our paper also relates to other studies on similar issues. Rodrik (2008) focuses on a related (but
markedly different) concept, that of undervaluation (and overvaluation). He emphasizes that
“Avoiding overvaluation of the currency is one of the most robust imperatives that can be
gleaned from the diverse experience with economic growth around the world, and it is one that
appears to be strongly supported by cross-country statistical evidence”, referring to the work of
Razin and Collins (1997), Johnson, Ostry, and Subramanian (2007), and Easterly (2005). We
differ from Rodrik as we focus on large appreciations and do not refer to a particular benchmark,
therefore abstracting from the question of over- or undervaluation. Our approach is motivated by
the considerable uncertainty that surrounds estimates of equilibrium exchange rates -- see for
instance the discussion in Bussière et al. (2010).5
The rest of the paper is organised as follows. Section 2 introduces the definition of large
exchange rate appreciations, and presents key stylized facts by means of event case analysis.
Section 3 presents a simple theoretical model and contrasts the growth pattern in appreciation
5 A given appreciation does not necessarily coincide with an overvaluation: it could be that the exchange rate is
converging towards a new equilibrium, correcting a past undervaluation. Having said that, the definition we use
excludes large appreciations that followed a currency crisis: given that such crises are well-known to give rise to an
overshooting effect, such episodes would most likely correspond to a correction towards an equilibrium value.
5
episodes depending on the nature of the underlying shock. Section 4 concludes and presents
possible policy implications.
2. The impact of large appreciations: stylized facts
2.1. Definition of large real exchange rate appreciations episodes
Our focus being on episodes of large and sustained appreciations, and not on temporary ones, we
first define the criteria used to identify these episodes in the data. Because of the limited number
of studies of large appreciations, we cannot rely on a large body of literature to provide a widely
accepted rigorous (numerical) definition of what constitutes a “large real appreciation”. Our
selection reflects two aspects. First, as we are interested in the aggregate macroeconomic
outcome, we focus on effective (rather than bilateral) exchange rates. Second, we establish clear
thresholds (for the nominal and real exchange rates) and take account of the behavior of the
exchange rate before and after the year of the appreciation to focus on lasting episodes and not
just sudden (and transitory) jumps.
We follow Kappler et al. (2011)’s definition of large appreciations, which requires the
simultaneous meeting of the following three criteria:
A. the nominal effective exchange rate is revalued by at least 10 percent or more relative to
the average level two years before, allowing to capture one-time step revaluations, but
also a number of smaller appreciation steps that happen within a short time window.
1.0)]ln[()]ln[( 2 ≥− −tt NEERNEER
B. the nominal appreciation must lead to sustained real appreciation, hence the real effective
exchange rate must remain stronger by, at least 10 percent on average for three years
relative to the beginning of the appreciation process.
[ ] [ ] [ ] [ ] 1.0)(ln3
)(ln)(ln)(ln2
321 ≥−++
−+++
t
ttt REERREERREERREER
C. There was no depreciation of similar magnitude prior to the appreciation
[ ] [ ] [ ] [ ]1.0
3
)(ln)(ln)(ln)(ln 543
1 −≥++
− −−−−
ttt
t
REERREERREERREER
6
These criteria ensure that the identified episode display a large movement in the exchange rate
(criteria A), that this movement is not reverted in the short run (criteria B) and that it does not
constitute a “catch up” following a large depreciation of the currency (criteria C).
Table 1 presents the identified episodes. The first column shows the ones that only meet criteria
A, i.e. the episodes where the currency underwent an appreciation of 10 percent or more in a
given year. We find 95 such country/time episodes. The second column shows the episodes that
meet all three criteria. Including all criteria narrows the number of large appreciations episodes
to 30, clearly illustrating how one needs to look beyond the behavior of the nominal exchange
rate in a given year. Out of these 30 episodes, 21 took place in advanced economies. The
inclusion of criteria B and C eliminates episodes where a large nominal appreciation is offset by
low domestic inflation (or deflation). For instance, all three large nominal appreciations by
Germany and five out of seven large nominal appreciations by Japan are excluded from the end
definition. Our definition also excludes a number of appreciations that took place just after a
currency crisis; this removes for example Korea and Thailand just after the Asian crisis. Among
the large advanced economies, it is noteworthy that the definition selects two episodes for the
United States: in the early 1980s (1982-83) and in the late 1990s (1997-98).
2.2. Variables of interest
We next turn to the main variables of interest in our analysis. As concerns about large
appreciations stress their detrimental effect through international trade, we consider overall GDP
growth, as well as the growth of exports and imports. All three variables are expressed in year-
on-year growth rates in real terms (first log differences).
In addition to looking at the unconditional effect of exchange rate appreciations on real output
and trade, we contrast the pattern between appreciations that are associated with unusual
developments in the financial sector with that of other appreciations. Financial market conditions
are proxied by large net capital inflows and increases in lending to the domestic sector. This
focus is motivated by the fact that swings in international financial markets are a major driver of
economic performance in emerging countries, but also in advanced ones as the current crisis
highlights.
7
Our definition of the “capital flow bonanza” follows Reinhart and Reinhart (2008). Inflows are
measured as a percentage of GDP, and correspond to the sum of portfolio and FDI inflows.6
Following Reinhart and Reinhart (2008), we construct a “capital flow bonanza” dummy, CFB,
which is equal to 1 in years when net inflows exceed the 80th percentile of the entire sample.7
Domestic credit to the private sector is also measured as a percentage of GDP. We follow the
definition of the World Bank and focus on domestic credit provided by the banking sector
including all credit to various sectors on a gross basis, with the exception of credit to the central
government, which is net.8 We construct a “lending boom” dummy variable that is equal to 1 in
years when domestic credit to the private sector exceeds the 80th percentile of the entire sample
for each country.
2.3. Descriptive statistics and event case analysis
We undertake an event case analysis that allows us to highlight key patterns in the relation
between large appreciations and output growth. The top panel of Table 2 presents the values of
GDP growth (mean and median) in episodes of large appreciations, both overall and
distinguishing between advanced and emerging countries. The bottom panel reports the
corresponding numbers in the absence of a large appreciation. Three key results stand out. First,
international trade behaves as expected. Episodes of large currency appreciations are associated
with weaker export growth: when all three criteria are met (i.e the dummy variable ABC is one),
the annual growth rate of real exports reaches 4.7% on average, compared to 6.2% otherwise.
This is in line with the competitiveness effect of the appreciation on exports. We note that this
effect is particularly strong for the advanced economies (3.4% against 5.7%), whereas for
6 According to the World Bank definition, portfolio equity includes net inflows from equity securities other than
those recorded as direct investment and including shares, stocks, depository receipts (American or global), and
direct purchases of shares in local stock markets by foreign investors; foreign direct investment are the net inflows
of investment to acquire a lasting management interest (10 percent or more of voting stock) in an enterprise
operating in an economy other than that of the investor. It is the sum of equity capital, reinvestment of earnings,
other long-term capital, and short-term capital as shown in the balance of payments. This series shows net inflows
(new investment inflows less disinvestment) in the reporting economy from foreign investors. 7 Here two definitions can be used: either consider capital flows in current terms and use the distribution for each
country (which leads to select the same number of episodes per country), or consider the overall distribution (i.e. for
the whole sample), expressed as a percentage of GDP to be able to compare small and large countries. We chose to
focus on the latter, although we also considered the former, with qualitatively similar results. 8 In the World Bank definition, the banking sector includes monetary authorities and deposit money banks, as well
as other banking institutions where data are available (including institutions that do not accept transferable deposits
but do incur such liabilities as time and savings deposits).
8
emerging market economies the difference is not very large. Similarly, large appreciations are
associated with stronger growth in imports (8.2% on average, against 6% in normal times, this
effect being particularly large in EMEs, 12.7% against 6.8%). This first set of results supports
the concerns expressed on the adverse competitiveness effect of large appreciations on net trade.
More surprisingly, our second result shows that the pattern for net exports is a poor proxy for
overall growth. In spite of the ceteris paribus effect described above, growth tends to be larger
during appreciation episodes (4.1%) than during normal times (3.3%). This is the case both for
advanced and emerging market economies, and particularly so for emerging economies. There
thus appears to be other factors, coinciding with large appreciations, which boost growth
sufficiently strongly to offset the adverse effect of the appreciation through trade. In other words,
it is important to bear in mind that the exchange rate and growth are both endogenous variables
that react to some underlying shock, and thinking solely of the impact of the exchange rate on
competitiveness can be highly misleading.
As Table 2 only shows static results with the contemporaneous growth rate of key variables
depending of whether the country experiences a large appreciation or not, we now turn to the
dynamic effects of appreciations. We use an event case analysis that encompasses not just the
contemporaneous impact but also its aftermath. Figure 1 shows the impact of a large exchange
rate appreciation by looking at the dynamic behavior on GDP growth (panel A), real imports
(panel B), and real exports (panel C). In each panel year t corresponds to the year when the three
criteria are met (ABC=1). Each panel shows the patterns across all countries, as well as among
advanced economies, and among emerging markets. Figure 1 also shows the path of the real
effective exchange rate itself (panel D) to illustrate its evolution after the time when the three
criteria are met. The variables in all panels are expressed in terms of the net effect relative to
normal times, i.e. when ABC=0. For example, the value at time t in panel A corresponds to the
difference shown in Table 2 between the average growth rate when ABC=1 (4.1%) and when
ABC=0 (3.3%), i.e. 0.8% for all countries taken together.
Figure 1 shows a substantial degree of heterogeneity. Panel A reveals that although countries
experiencing a large appreciation grow more robustly on average at the time of the depreciation,
the effect varies over time and across country groups. Advanced economies experience a
sustained pickup in growth, while emerging economies only experience a temporary increase in
9
growth, followed by a contraction within two years. In line with this markedly different
evolution of aggregate demand, import growth tends to accelerate after the appreciation for
advanced economies, but strongly decelerates for emerging economies (panel B). Turning to
export growth (panel C), a large appreciation reduces exports on impact for advanced economies.
This effect persists for one year, after which exports recover. Emerging countries, by contrast,
see a better performance of exports on impact, but this is temporary and it is followed by a
sizable deterioration. Figure 1 therefore reveals that the response of key macroeconomic
variables differs dramatically across countries, following a large appreciation. Among the
different factors that may explain this, we focus on the role of large capital inflows and domestic
lending. Our focus on these proxies for conditions in financial markets is motivated by the well-
documented role of financial shocks in driving macroeconomic conditions in emerging markets,
but also in advanced economies.
Figure 2 shows how the impact of large appreciations differs depending on whether the economy
also experiences a capital flow bonanza (top panels) or a lending boom (bottom panels). All
panels show the dynamic pattern on GDP growth at the time of the large appreciations and in
subsequent years. In addition, each panel presents the growth difference between large
appreciation episodes (ABC=1) and normal times, a positive value indicating higher growth
during a large appreciation episode.
The impact of a large appreciation is quite sensitive to the simultaneous presence of a capital
flows bonanza. Panel E shows that an appreciation occurring along with a bonanza is associated
with a moderate increase in growth in advanced economies, but a contraction in emerging
economies. By contrast, a large appreciation that is not accompanied by a bonanza (Panel F) is
associated with somewhat stronger growth in advanced economies, and a sizable – albeit short
lived – boom in emerging markets. Considering the role of credit booms instead of bonanzas
leads to a similar picture. Appreciations that are concurrent to credit booms are associated with
sharp contractions in emerging economies (Panel G), whereas countries experience a temporary
acceleration in activity when the appreciation is not accompanied by a credit boom (Panel H).
The pattern is similar for advanced economies, albeit with smaller magnitudes. One potential
reason why results differ across country groups may be related to the fact that financial
supervision is in general more developed in advanced countries, compared to developing
10
countries, such that “capital flow bonanzas” and “lending booms” may be more disruptive in the
latter.
To sum up, our analysis shows that while large appreciations have the expected adverse effect on
net trade, this does not lead to an overall contraction in GDP, to the contrary. The event study
analysis shows a substantial extent of heterogeneity through time, across countries, and
depending on the underlying financial conditions. Most strikingly, large appreciations associated
with capital flow bonanzas or lending booms are characterized by a weaker growth performance
than appreciations occurring without bonanzas or booms.
3. The contrasted impact of productivity and capital flows shocks: a simple model
In this section we present a simple model that contrasts the impact of productivity shocks
with that of financial shocks on the real exchange rate, output and lending. For brevity we focus
on the main features and results, and leave more details to the appendix.9 We first present the
building blocks and the solution method. We then derive the analytical solution for a simple
combination of parameters, and present numerical results for the more general parametrizations.
3.1 Building blocks
Our setup builds on Obstfeld and Rogoff (1996 ch 4). We consider a general equilibrium
model with two countries, Home and Foreign, of sizes n and 1-n respectively. In the Home
country, a representative agent of size n consumes a basket tC of non-traded and traded goods,
with the latter consisting of Home and Foreign goods:
( ) ( ) ( )
( ) ( ) ( ) ( ) 11
,
11
,
1
,
11
,
11
,
1
)1)(1()1(
1
−−−
−−−
−−+−+=
−+=
θθ
θθ
θθθ
θ
λλ
λλ
λλλ
λ
χχ
γγ
tFtHtT
tNtTt
CnCnnC
CCC
9 The fully detailed steps of the model solution, and associated Matlab programs, are available on request.
11
where t denotes time, tTC , and tNC , are the consumptions of traded and non-traded goods,
respectively, and tHC , and tFC , are the consumptions of Home traded and Foreign traded goods,
respectively. λ is the elasticity of substitution between traded and non-traded goods, and θ is the
elasticity of substitution between Home traded and Foreign traded goods. [ ]1,0∈χ is the degree
of domestic bias in traded goods consumption.
The consumption baskets of the representative agent of size 1-n are similar, with asterisks
denoting Foreign variables:
( ) ( ) ( )
( ) ( ) ( ) ( ) 11*
,
11*
,
1*
,
11*
,
11*
,
1
*
1)1(
1
−−−
−−−
+−+−=
−+=
θθ
θθ
θθθ
θ
λλ
λλ
λλλ
λ
χχ
γγ
tFtHtT
tNtTt
CnnCnC
CCC
The allocation of consumption reflects the various relative prices, namely the price of
Home traded goods tHP , the price of Foreign traded goods tFP , (both prices are the same in the
two countries), the price indexes of traded goods tTP , and *
,tTP , and the consumer price indexes
tP and *
tP . The price indexes are:
( )( ) ( )[ ]( ) ( )( )[ ]
( ) ( )( )[ ] ( ) ( )( )[ ] λλλλλλ
θθθ
θθθ
γγγγ
χχ
χχ
−−−−−−
−−−
−−−
−+=−+=
+−+−=
−−+−+=
1
11*
,
1*
,
*1
11
,
1
,
1
11
,
1
,
*
,
1
11
,
1
,,
1 ; 1
1)1(
)1)(1()1(
tNtTttNtTt
tFtHtT
tFtHtT
PPPPPP
PnnPnP
PnPnnP
We define the terms-of-trade as the price of Foreign traded goods to Home traded goods,
tHtFt PPT ,, /= , and the relative prices of non-traded goods relative to traded goods as
tTtNt PPR ,, /= and *
,
*
,
* / tTtNt PPR = . The real exchange rate ttt PPQ /*= reflects the terms-of-trade
(in the presence of domestic bias) and the relative prices of non-traded goods:
12
( )( )( ) ( )
( )( )( )( )
λ
λ
λθ
θ
θ
γγγγ
χχχχ −
−
−−
−
−
−+−+
−−+−++−+−
=1
1
1
1*1
1
1
1
1
1
)1)(1()1(
1)1(
t
t
t
t
tR
R
Tnnn
TnnnQ
The production of traded and non-traded goods relies on a technology that uses labor with
decreasing returns to scale. The total labor supply in the Home and Foreign country are set to n
and 1-n. The outputs of the two sectors in the Home country are α−−= 1
,,, )( tNtHtH LnAY and
α−= 1
,,, )( tNtNtN LAY where tNL , denotes the labor input in the non-traded sector, and tiA , is an
exogenous productivity term in sector i = H, N that is our first source of shocks. The parameter
[ ]1,0∈α reflects the degree of returns to scale. The case of α = 1 corresponds to an endowment
economy, while the case of α = 0 corresponds to constant returns to scale. The outputs in the
Foreign country are α−−−= 1*
,
*
,
*
, )1( tNtFtF LnAY and α−= 1*
,
*
,
*
, )( tNtNtN LAY .
Borrowing and lending takes place through a bond denominated in Foreign traded goods,
without loss of generality. A unit of bond held between period t and t+1 yields an interest rate of
11 ++ tr . We denote the per capita holdings of bonds by the Home agent at the end of period t by
1+tB . The intertemporal constraints faced by the Home and Foreign agent are:
ttFttNtNtNtNtFtFttFtt
ttFttNtNtNtNtHtHttFtt
BPrnLAPLnAPBnPCPn
BPrnLAPLnAPBnPCnP
,
1*
,
*
,
*
,
1*
,
*
,,1,
**
,
1
,,,
1
,,,1,
)1()()()1(
)1()()(
+−+−=−−
+++−=+−−
+
−−+
αα
αα
where we used the fact that bonds are in zero net supply worldwide. As in each country the
consumption of non-traded goods is equal to its supply, we split the constraint in each country
between the market clearing condition for the non-traded sector:
αα −− =−= 1*
,
*
,
*
,
1
,,, )()1( ; )( tNtNtNtNtNtN LACnLAnC (1)
and the intertemporal constraint in terms of traded goods:
( ) ttttNtHtttFttH BTrnLnABnTCTCn )1()( 1
,,1,, ++−=++ −+
α (2)
( ) ttttNtFttttFttH BTrnLnATBnTCTCn )1()()1( 1*
,
*
,1
*
,
*
, +−−=−+− −+
α (3)
13
The clearing of the market for Home traded goods requires that the supply matches the
demand:
α−−=−+ 1
,,
*
,, )()1( tNtHtHtH LnACnnC (4)
A similar relation holds for the clearing of the Foreign traded good, but is redundant given (1)-
(4). The appendix presents the expressions for (1)-(4) using the expressions for the intratemporal
allocation of consumption.
The Home representative agent maximizes an intertemporal utility of consumption over
an infinite horizon ( ) ( )∑∞
=++=
0
, lns
st
s
stHt CU β . The Foreign representative maximizes a similar
utility ( ) ( )*
0
,
* ln st
s
s
stFt CU +
∞
=+∑= β . We assume that the discount factors of the Home agent, stH +,β ,
can differ from that of the foreign agent, stF +,β . These discount factors are our second source of
shock. An increase in the Foreign factor represents a higher willingness to lend by the Foreign
agent, leading to capital flows towards the Home country all else equal.
The intertemporal optimization leads to two conditions for each country. The first ones
are the Euler conditions for the dynamics of consumption:
**
11,
*
111,1 )1( ; )1( t
C
ttFtt
C
ttHt CrCCrC ++++++ +=+= ββ (5)
where C
tr 1+ and *
1
C
tr + are the real interest rates in terms of the consumption baskets:
*
1
*
,
1,
1
*
1
1,
1,
11 )1(1 ; )1(1+
+++
+
+++ +=++=+
t
t
tF
tF
t
C
t
t
t
tF
tF
t
C
tP
P
P
Prr
P
P
P
Prr (6)
The second set of conditions reflects the optimal allocation of labor across the traded and non-
traded sectors:
14
αα −− =− )()( ,,,,,, tNtNtNtNtHtH LAPLnPA (7)
αα −− =−− )()1( *
,
*
,
*
,
*
,,
*
, tNtNtNtNtFtF LAPLnPA (8)
The appendix presents the real interest rates in (5) and the allocations (6)-(7) in terms of relative
prices. Note that when Home and Foreign traded goods are perfect substitutes, so that
*
,,,, tFtTtHtH PPPP === , and there are constant returns to scale in production, (7)-(8) imply that
the relative price of non-traded goods only reflects relative productivities: tNtHt AAR ,, /= and
*
,
*
,
* / tNtFt AAR = .
3.2 Solution method
As the model is highly non-linear, we approximate it around a symmetric steady state. In
that steady state, where variables are indexed by 0, agents are equally patient ( 00,0, βββ == FH )
which removes incentives to save and borrow internationally. In addition no country holds
claims on the other ( 00 =B ). The Euler conditions (5) imply that the real interest rates are all
equal to the inverse of the discount rate.
The baseline steady-state is characterized by the two market clearing conditions for non-
traded goods (1), the intertemporal constraints (2)-(3), the market clearing for the Home traded
good (4), and the labor allocations (7)-(8). For simplicity, we put restrictions on the productivity
levels10
that ensure that all relative prices are unity ( 1*
000 === RRT ), that consumption levels
are equalized in the two countries (αγ −== )(0,
*
00 nACC H ), and that the labor allocations reflect
the weight of the two sectors in preferences ( )1(0, γ−= nLN and )1)(1(*
0, γ−−= nLN ).
We express the model in terms of log-linear approximations around the baseline steady
state and denote log deviations by hatted values, with for instance 00 /)(ˆ CCCC tt −= . As
10 Specifically, we set
αγγ −−= ))1/((0,0, HN AA , α−−= ))1/((0,
*
0, nnAA HF and
ααγγ −− −−= ))1/(())1/((0,
*
0, nnAA HN .
15
international bond holdings are zero in the steady state, we define )/(ˆ0CBB tt γ= . The
approximations of equations (1)-(8) are presented in the appendix.
We consider that the economy is initially in the baseline steady state. The economy is then
hit by shocks, and we distinguish between the short run (first period with shocks) and the long
run (subsequent periods). Shocks in the short run affect the discount factors, Hβ and Fβ , and
the productivity levels HA , NA , *ˆFA and
*ˆNA . In the long run the economy reaches a new
steady state where the discount factors revert to their initial value 0β and productivity levels
stabilize at levels that can differ from the initial ones. The long run productivity levels are
denoted by ssHA ,ˆ , ssNA ,
ˆ , *
,ˆ
ssFA and *
,ˆ
ssNA with the ss subscript denoting the new steady state.
With this specification, the model is a linear system with 18 endogenous variables,11
10
exogenous variables,12 and 18 equations.13 In addition to the 18 variables, we compute the real
exchange rate of the Home country that reflects the terms-of-trade (to the extent that there is
domestic bias in consumption) and the cross-country difference in the relative price of non-
traded goods:
)ˆˆ)(1(ˆˆ *
tttt RRTQ −−−= γχ (9)
The output in the Home sectors and the overall Home output, measured in terms of Home traded
goods, are written as:
11
The Home and Foreign consumptions in the short and long run, the relative prices of non-traded goods in the
Home and Foreign countries in the short and long run, the terms-of-trade in the short and long run, the labor used in
the non-traded sectors in the Home and Foreign countries in the short and long run, the Home assets accumulated in
the short run, the real interest rates in terms of consumption baskets between the short and long run, and the real
interest rate in terms of the Foreign traded goods between the short and long run.
12 The shocks to productivity levels in the four sectors (Home and Foreign traded and non-traded) in the short and
long run and the shocks to the Home and Foreign discount factors between the short and long run.
13 The first four equations are the market clearing conditions for non-traded goods (1) in the short and long run, the
next four are the intertemporal constraints (2)-(3) in the short and long run, the next two are the market clearing for
the Home traded good (4) in the short and long run, the next two are the Euler equations (5) between the short and
long run, the next two are the real interest rates in terms of consumption baskets (6) between the short and long run,
and the final four are and the labor allocations (7)-(8) in the short and long run.
16
[ ]tttNtHt
tNtNtNtNtHtH
TnRAAY
LAYLAY
ˆ)1)(1(ˆ)1(ˆ)1(ˆˆ
ˆ)1(ˆˆ ; ˆ1)1(ˆˆ
,,
,,,,,,
χγγγ
αγ
γα
−−+−+−+=
−−=−−−= (10)
While this linear system has a unique solution, the analytical solution is quite
cumbersome for the general model. We therefore first present the analytical solution for a
specific case, before presenting a numerical illustration of the general case.
3.3 A simple case
We simplify the model in two ways. First, we consider that Home and Foreign goods are
perfect substitutes ( ∞→θ ), so that there is a common traded good. The terms-of-trade are then
always equal to one, so we lose one endogenous variable as 0ˆ =tT . We also lose one equation as
the market clearing condition (4) corresponds to the sum of (2) and (3). In addition the degree of
domestic bias χ is irrelevant. The second simplification is to set the elasticity of substitution
between traded and non-traded goods λ equal to one.
In the long-run the economy reaches a new steady-state, with the cross-country asset
holdings ssB as a state variable. For brevity, we focus on short term shocks, and present the
solution in terms of cross-country differences. When the Home country is a net creditor ( 0ˆ >ssB
), it is characterized by higher consumption, relative to the Foreign one, a higher price of non-
traded goods, which translates into an appreciated currency, and a shift of labor towards the non-
traded sector:
n
BLLRRQ
n
BRR
n
BCC
ss
ssNssNssssss
ss
ssss
ss
ssss
−−
=−−−−=
−−
=−−
−=−
1
ˆ1ˆˆ ; )ˆˆ)(1(ˆ
1
ˆ1ˆˆ ; 1
ˆ1ˆˆ
0
0*
,,
*
0
0*
0
0*
ββγγ
ββα
ββγ
17
The cross-country asset holdings ssB are endogenously determined in the short run. We
define a summary intertemporal measure of shocks )ˆˆ()ˆˆ(ˆ *
FHFH AAZ −+−= ββ . A positive value
indicates that the Home country is relatively more patient ( FH ββ ˆˆ > ) or that it has a temporary
productivity advantage in the traded sector ( 0ˆˆ * >− FH AA ). The Home country accumulates
assets if this measure is positive:
Zn
Bss ˆ)1(1
ˆ0
αγγβ−+
=−
Short run consumption is higher in the Home country, relative to the Foreign country, if it
benefits from higher productivity or if the intertemporal pattern of shocks leads it to borrow (
0ˆ <Z ). The Home country is characterized by a higher price of non-traded goods, and an
appreciated real exchange rate, if it has an advantage in traded sector productivity or if the
intertemporal pattern of shocks leads it to borrow, thereby boosting demand for non-traded
goods:
)ˆˆ)(1(ˆ
ˆ)1(
)ˆˆ()ˆˆ(ˆˆ
ˆ)1(
)ˆˆ)(1()ˆˆ(ˆˆ
*
0***
0***
RRQ
ZAAAARR
ZAAAACC
ss
NNFH
NNFH
−−−=
−+−−−−=−
−+−−−+−=−
γαγγ
βα
αγγβγγγ
Borrowing by the Home country shifts the allocation of labor towards the non-traded
sector. The Home country is finally characterized by a high real interest rate (in terms of
consumption basket) if the intertemporal pattern of shocks leads it to borrow, or if it has an
advantage in traded sector productivity:
[ ])ˆˆ()ˆˆ()1(ˆ)1(
)1()(
ˆ)1(
ˆˆ
***
0
0*
,,
NNFH
CC
ssNssN
AAAAZdrdr
ZLL
−−−−+−+
−−=−
−+−=−
γαγγ
αγβ
αγγβγ
18
We illustrate the impact of the various shocks through a numerical example. We set the
Home country to be very small with n =0.1. We assume that traded good account for 30 % of the
total consumption basket (γ = 0.3), that there are decreasing returns to scale (α = 0.3), and set the
discount factor β0 to 0.95.
Figure 3 presents the impact of three shocks on the short-run Home variables. These are a
temporary increase in Home traded productivity ( 0ˆ >HA , grey bars), a permanent increase in
Home productivity ( 0ˆˆ, >= ssHH AA , stripped bars), and an increase in Foreign patience ( 0ˆ >Fβ
black bars). The increase in Foreign patience can be interpreted as a capital inflows shock.
Another approach to assess the impact of financial conditions on the exchange rate is found in
Benigno and Romei (2012) who consider a tightening of borrowing constraints, but abstract from
the distinction between traded and non-traded goods. To facilitate the comparison across the
three shocks, they are calibrated to lead to a unit appreciation in the Home currency.
The responses of the various variables depend crucially on the nature of the shock.
Consumption increases most when the real appreciation reflects a capital inflow, and is financed
by running a sizable current account deficit. By contrast a temporary productivity shock is
smoothed through a current account surplus and leads to a contraction in consumption. All three
shocks raise the price of non-traded goods.14
The impact on the real interest rate (in terms of the
consumption basket) is contrasted: a temporary productivity improvement raises the interest rate
and induces a postponement of consumption, whereas a surge in capital inflows lowers the
interest rate and brings consumption forward.
Overall output increases under all shocks, but does so substantially more under a
productivity improvement. The composition of output markedly differs across shocks. An
increase in productivity in the traded sector leads to a shift of labor towards that sector. As a
result traded output surges thanks to more labor used in that sector and higher productivity, while
non-traded output falls because of the labor reallocation. The pattern is opposite under a shock to
14
One may notice that the relative price of non-traded goods in the Home country does not increase equally across
all shocks, even though the real exchange rate impact is the same. This simply reflects the fact that the real exchange
rate is also affected by the relative price of non-traded goods in the Foreign country.
19
patience. The capital inflows into the Home country fuel a higher demand for non-traded goods
that is met by reallocating labor through the traded sector. As a result, traded output falls.
Figure 3 thus clearly illustrates the central result of the model, namely that a given
appreciation of the real exchange rate can be associated with sharply different situations. An
appreciation due to a productivity gain is associated with an output boom led by the traded
sector, a higher real interest rate, and a current account surplus (if the gain is temporary). An
appreciation fuelled by capital inflows by contrast leads to a credit-fuelled consumption boom,
with a low real interest rate, and only a moderate overall output boom as the surge of non-traded
output is offset by a contraction in traded output.
3.4 The general case
We now turn to the general case where the elasticities of substitution θ and λ can take
more general values. As the analytical solution is complex, we focus on a numerical illustration.
We keep the same parameters as in the previous section and assume that there is no domestic
bias in consumption (χ = 0).
Figures 4 to 5 show the response of the short run Home variables to a unit temporary
improvement in productivity in the traded sector ( 1ˆ =HA , Figure 3), a permanent improvement (
1ˆˆ, == ssHH AA , Figure 4) and a temporary increase in Foreign patience ( 1ˆ =Fβ , Figure 6). Each
panel shows the value of the variables for different elasticities of substitutions between traded
and non-traded goods (λ going from 1 to 8 on the horizontal axis) and between Home and
Foreign traded goods (θ being set at 1, 3, 6 or infinity). The particular case of the previous
section corresponds to the leftmost point of the thick line. For brevity we focus our discussion on
the main patterns.
Allowing for more elasticity between traded and non-traded goods, or allowing for a
limited substitutability beteween Home and Foreign traded goods, reduces the magnitudes of the
impact of shocks.
A temporary improvement of traded productivity (Figure 4) lowers consumption and
raises savings, unless Home and Foreign traded goods are poor substitutes in which case
consumption increases. The higher productivity reduces the price of Home traded goods and
20
deteriorates the terms-of-trade ( 0ˆ >T ) especially when traded goods are poor substitutes among
themselves. The relative price of non-traded goods increases, leading to a real appreciation,
although the magnitude depends on the elasticities. The impact on traded and non-traded outputs
and labor reallocation is qualitatively similar to the simple case, but is dampened by a low
substitutability between Home and Foreign traded goods. The impact on the elasticities of
substitution is broadly similar when the improvement in productivity is permanent (Figure 5).
The sensitivity of the various variables to the elasticities of substitution is more moderate
when the economy is hit by a shock to Foreign patience (Figure 6). The magnitude of the
consumption boom and borrowing is somewhat higher when there is more substitutability
between traded and non-traded goods, or between Home and Foreign traded goods. The
movements in the relative price of non-traded goods and the real exchange rate are lowered when
traded and non-traded goods are closer substitute, as any reallocation then requires less
movement in prices. The impact of the shocks on traded and non-traded outputs and labor
reallocation is qualitatively similar to the simple case, but is dampened by a low substitutability
between Home and Foreign traded goods or by a high substitutability between traded and non-
traded goods. As labor is reallocated towards the non-traded sector, the output of Home traded
goods falls. This raises their price and leads to an improvement in the terms-of-trade, especially
when a limited substitutability between Home and Foreign traded goods limits the impact of
price movements on quantities.
To facilitate the contrast between the various shocks, Figure 7 presents their impact when
the shocks are calibrated to lead to a unit real appreciation of the Home currency in the short run.
For each variable we show the impact of a temporary increase in productivity in the traded sector
(left panel), that of a permanent increase (middle panel) and that of a capital inflows stemming
from higher patience in the Foreign country (right panel).15
The impact of consumption is relatively insensitive to the elasticity of substitution
between Home and Foreign traded goods, but react more to the elasticity of substitution between
traded and non-traded goods, with a higher elasticity leading to a larger increase in consumption.
The panels show that the consumption boom is much larger in the case of a capital inflows
15
The rounded line corresponds to θ = 1 in Figures 2 to 4, but to θ = 1.5 in Figure 5 as the real exchange rate
movement can be zero when θ = 1.
21
shocks than in the case of a productivity shock. A higher elasticity of substitution between traded
and non-traded goods raises the magnitude of the current account response to temporary shocks.
The terms-of-trade react most when Home and Foreign traded goods are poor substitutes,
as a given quantity adjustment then requires larger price movements. The contrast between a
productivity improvement, which worsens the terms-of-trade, and a capital inflows boom, which
improves them, is larger when traded and non-traded goods are poor substitutes. The relative
price of non-traded goods by contrast shows limited sensitivity to the elasticities, which simply
reflects our parametrization that the real exchange rate appreciates by the same value in all cases.
The pattern for the real interest rate is robust to the elasticities of substitution, with a
productivity improvement leading to a higher interest rate (unless Home and Foreign traded
goods are poor substitutes and traded and non-traded goods are close substitutes) and a capital
inflows increase lowering the interest rate, especially with higher substitutability between traded
and non-traded goods.
Moving away from the simple case by increasing the substitutability between traded and
non-traded goods or lowering that between Home and Foreign traded goods magnifies the
impacts on outputs. A productivity improvement substantially raises traded output, at the
expense of non-traded output, with the opposite pattern being observed following an increase in
Foreign patience. Moving away from the simple model thus reinforces the contrast shown in
Figure 3.
Our analysis so far abstracts from domestic bias in consumption, which implies that the
real exchange rate only reflects the relative prices of non-traded goods. We relax this restriction
by setting χ = 0.3, which implies that the ration between the demand for local and imported
traded goods is three times as large as in the absence of domestic bias. Figures 8 to 10 show the
impact of temporary and permanent improvements in traded productivity and the impact of a
temporary increase in Foreign patience, and correspond to Figures 4 to 6 where χ was set to zero.
For brevity we focus on the aspects that are affected by domestic bias in consumption.
Following an improvement in Home traded productivity, the most salient difference is for the
real exchange rate, as it is now pulled in two opposite directions. First, the higher traded
productivity lowers the price of Home traded goods, which raises the relative price of non-traded
goods, and tends to appreciate the real exchange rate as before. The reduction in Home traded
price also worsens the term-of-trade, which now pushes the real exchange rate towards a
22
depreciation. The balance between these two effects depends on the elasticities of substitution,
and a real depreciation occurs when Home and Foreign traded goods are poor substitutes (as the
terms-of-trade then move by more) and traded and non-traded goods are closer substitutes (as the
relative price between them then moves by less).
The real interest rate is also affected, and increases by less than in the absence of domestic
bias. As a result, Home consumption raises, or falls by less than in the absence of bias. The
pattern of the various outputs is little affected by the presence of domestic bias.
Turning to the impact of a permanent increase in Home traded productivity shows a
similar pattern. The impact of domestic bias in consumption is however limited, with a few
exception such as the real exchange rate when Home and Foreign traded goods are poor
substitutes.
The presence of domestic bias in consumption affects the response of the variables to a
shock in Foreign patience to some extent, primarily when there is little substitutability between
Home and Foreign traded goods. In that case, the consumption boom somewhat reduced, and the
real exchange rate appreciation magnified as the improvement in the terms-of-trade now
reinforces the impact of the higher relative price of non-traded goods. The impact on the various
outputs is also somewhat reduced.16
Overall our numerical illustration of the general model reinforces the patterns shown in
Figure 3. A real appreciation due to a productivity improvement in the traded sector leads to an
output boom in that sector and a current account surplus. The same appreciation stemming from
an increase in Foreign savings is associated with a credit fuelled consumption boom and a
limited increase in overall output as resources shift towards the non-traded sector.
4. Conclusion
This paper assesses the effect of large currency appreciations on growth and trade, by means of
stylized facts and event case analysis, and shows how the identified patterns are consistent with a
simple theoretical model. The main empirical results are threefold. First, and as expected, large
exchange rate appreciations are associated with weaker export growth and stronger import
16
As in some cases the real exchange rate depreciates, we do not build a figure corresponding to Figure 5 where a
real unit appreiciaton is imposed.
23
growth, compared to normal times. However, the second result is that output growth is actually
stronger in the wake of a large appreciation, suggesting that other factors are sufficiently
powerful to offset the effect through trade. Third, there is a substantial degree of heteorogeneity
across appreciation episodes. Appreciations that are associated with so-called capital flow
bonanzas (large capital inflows) or “lending booms” (large increases in lending to the private
sector) are characterized by weaker growth compared to episodes with no lending boom and no
capital flow bonanzas. These patterns are consistent with a simple theoretical model which
shows that appreciations stemming from productivity improvements are associated with a
stronger economic performance than appreciations stemming from easier borrowing conditions
in world financial markets.
In terms of policy implications, our analysis shows that policy should not be designed solely in
response to exchange rate movements, but instead needs to identify the driving factors. Financial
inflows and credit booms emerge as sources of concern. These however are likely to be best
dealt with through targeted management of capital flows and credit growth, for instance using
macroprudential tools. A policy aimed at the exchange rate, which is merely a consequence of
the underlying shocks, could well be too blunt a tool to effectively address legitimate policy
concerns.
24
REFERENCES
Benigno, P., and F. Romei (2012), “Debt Deleveraging and the Exchange Rate”, NBER Working Paper
No. 17944.
Berg, A. and C. Pattillo (1999), “Predicting Currency Crises: The Indicators Approach and an
Alternative”, Journal of International Money and Finance; 18(4), August 1999, pp. 561-86.
Bussière, M. and M. Fratzscher (2006), “Towards a New Early Warning System of Financial Crises”,
Journal of International Money and Finance, 25(6), October 2006, pp. 953-973.
Bussière, M., M. Ca’Zorzi, A. Chudik and A. Dieppe (2010), “Methodological Advances in the
Assessment of Currency Misalignments”, ECB Working Paper No. 1151, January 2010.
Bussière, M., S. Saxena and C. Tovar (2012), “Chronicle of Large Currency Devaluations: Re-Examining
the Effects on Output”, Journal of International Money and Finance, Volume 31, Issue 4, June 2012, pp.
680–708.
Calvo, G. A. and C. M. Reinhart (2002), “Fear Of Floating,” The Quarterly Journal of Economics, MIT
Press, vol. 117(2), pp. 379-408, May.
Cerra, V. and S. Saxena (2008), “Growth Dynamics: The Myth of Economic Recovery”, The American
Economic Review, Vol. 98, No. 1, pp 439-457.
Easterly, W. (2005), ”National policies and economic growth” in Philippe Aghion and Steven Durlauf,
editors, Handbook of Economic Growth, Elsevier, 2005.
Eichengreen, B., A. Rose and C. Wyplosz (1995), “Exchange Market Mayhem: The Antecedents and
Aftermaths of Speculative Attacks”, Economic Policy, 21, pp. 249-312.
Frankel, J. and A. Rose (1996), “Currency Crashes in emerging Markets: an Empirical Treatment”,
Journal of International Economics, 41, pp. 351-66.
Goldstein, M., and C. M. Reinhart (2000), “Assessing Financial Vulnerability: An Early Warning System
for Emerging Markets,” Peterson Institute Press: All Books, Peterson Institute for International
Economics, number 100, January.
Gourinchas, P.-O., R. Valdes and O. Landerretche. "Lending Booms : Latin America and the World."
Economia 1, 2 (2001).
Johnson, S. H., J. Ostry, and A. Subramanian (2007), "The Prospects for Sustained Growth in Africa:
Benchmarking the Constraints", March 2007, IMF Working Paper No. 07/52.
Kaminsky, G., S. Lizondo and C. M. Reinhart (1998), “Leading Indicators of Currency Crises,” IMF Staff
Papers, Palgrave Macmillan, vol. 45(1), pp. 1-48, March.
Kappler, M., H. Reisen, M. Schularick, and E. Turkisch, E. (2012), “The Macroeconomic Effects of
Large Exchange Rate Appreciations”, Open Economies Review, DOI 10.1007/s11079-012-9246-4.
Krugman, P. (1994), “Competitiveness: a Dangerous Obsession”, Foreign Affairs, Issue: March/April
1994 (volume 73, number 2).
25
Obstfeld, M., and K. Rogoff (1996). Foundation of International Macroeconomic, Princeton University
Press.
Razin, O. and S. M. Collins (1997), ”Real Exchange Rate Misalignments and Growth,” Georgetown
University, 1997.
Reinhart, C. M. and V. R. Reinhart (2008), “Capital Flow Bonanzas: An Encompassing View of the Past
and Present”, NBER Working Paper No. 14321, Issued in September 2008.
Rodrik, D. (2008), “The Real Exchange Rate and Economic Growth: Theory and Evidence”, Brookings
Papers on Economic Activity, Fall 2008, pp. 365-412.
26
Table 1: List of country episodes selected with the first criterion and in the final definition
Country Years identified according to: Country Years identified according to:
A A, B and C A A, B and C
Argentina 1993-2001 1997-1998 Korea 1989-1990 2005-2005
Australia 1973-1974 2004-2004 2000-2000
1981-1981 2005-2006
1989-1990
1997-1997
2003-2004 Malaysia 1983-1983
2011-2011 1993-1993
Brazil 2005-2008 2007-2008 Morocco 1989-1989
2011-2011 1994-1994
Netherlands 1987-1987
Canada 1989-1989 2004-2006 New Zealand 1994-1996 1995-1995
2004-2007 2003-2005 2004-2004
2011-2011 2011-2011
Chile 1981-1982 1992-1995 Norway 1974-1974
1989-1995 2002-2002
2005-2006 Pakistan 1981-1981
China 1975-1976 Peru 2001-2001
1980-1980 Philippines 2007-2007 2007-2007
1989-1989 Poland 2001-2001
1997-1998 2006-2006
2009-2009 2008-2008
Colombia 1994-1995 1995-1995 Saudi Arabia 1974-1974
2005-2005 2008-2008 1981-1983
2008-2008 1997-1998
2011-2011 Singapore 1982-1983 1982-1982
Czech Rep. 2002-2003 2002-2003 1990-1991 1990-1991
2006-2006 2006-2006 South Africa 2004-2004
2008-2008 2008-2008 2010-2010
Euro Area 1986-1987 1987-1987 Spain 1988-1990 1988-1989
2003-2004 2003-2004 Sweden 2011-2011
2008-2008 Switzerland 1974-1979 1977-1977
Finland 1995-1995 1983-1983
1987-1987
1995-1995
Germany 1973-1974 2010-2011
1977-1978 Thailand 1999-1999 2007-2007
1987-1987 UK 1980-1981 1980-1981
Hong Kong 1994-1994 1997-1998 1997-1998
Hungary 1976-1977
1980-1982 USA 1976-1977 1982-1983
Indonesia 2010-2010 1981-1985 1997-1998
Ireland 2003-2004 2003-2004 1989-1991
Israel 2008-2008 2008-2008 1993-1994
Italy 1997-1997 1997-1998
Japan 1972-1973 1984-1987 Uruguay 1981-1982 1993-1993
1977-1979 1992-1994 1989-1990
1981-1981 1993-1994
1984-1989 2006-2006
1992-1995 2011-2011
1999-2000
2009-2010
27
Table 2 : Summary statistics: impact of a strong appreciation
In situation of strong appreciation
GDP export import (ABC=1)
All countries
mean 4.1 4.7 8.2
median 4.1 4.0 8.3
Developed countries
mean 3.4 3.4 6.6
median 3.9 2.9 7.8
EMEs countries
mean 5.9 8.4 12.7
median 5.3 8.7 10.0
Not in situation of strong appreciation
GDP export import (ABC=0)
All countries
mean 3.3 6.2 6.0
median 3.3 6.1 6.2
Developed countries
mean 2.6 5.7 5.4
median 2.6 5.4 5.8
EMEs countries
mean 4.2 7.0 6.8
median 4.8 7.1 8.1
28
Figure 1: event case analysis
Impact of a strong appreciation at time t on
Panel A: Growth rate Panel B: Import growth
Panel C: Export growth Panel D: REER
-2,00
-1,50
-1,00
-0,50
0,00
0,50
1,00
1,50
2,00
t t+1 t+2 t+3
GDP_all GDP_devd GDP_EME
-6,00
-4,00
-2,00
0,00
2,00
4,00
6,00
8,00
t t+1 t+2 t+3
Imp_all Imp_devd Imp_EME
-3,00
-2,00
-1,00
0,00
1,00
2,00
3,00
t t+1 t+2 t+3
Exp_all Exp_devd Exp_EME-2,00
0,00
2,00
4,00
6,00
8,00
t t+1 t+2 t+3
REER_all REER_devd REER_EME
29
Figure 2: event case analysis
Impact of a Capital Flow Bonanza on
Panel E:Growth rate, CFB=1 at time t Panel F:Growth rate, CFB=0 at time t
Impact of a lending boom on
Panel G:Growth rate, LB=1 at time t Panel H:Growth rate, LB=0 at time t
-2,00
-1,00
0,00
1,00
2,00
3,00
4,00
5,00
t t+1 t+2 t+3
GDP_all GDP_devd GDP_EME-2,00
-1,00
0,00
1,00
2,00
3,00
t t+1 t+2 t+3
GDP_all GDP_devd GDP_EME
-10,00
-8,00
-6,00
-4,00
-2,00
0,00
2,00
4,00
t t+1 t+2 t+3
GDP_all GDP_devd GDP_EME
-1,00
-0,50
0,00
0,50
1,00
1,50
2,00
2,50
t t+1 t+2 t+3
GDP_all GDP_devd GDP_EME
30
Figure 3: Impact on short run Home variables, simple model
All shocks imply a unit Home real appreciation in the short run
Parameters: n = 0.1, l = 1, q = infinity, g = 0.3, a = 0.3, b0 = 0.95
-6 -4 -2 0 2 4 6 8
Consumption
Savings
Price of non-traded
Real exchange rate
Real int rate (cons units)
Output traded
Output non-traded
Output total (traded units)
Labor in non-traded
Temporary increase in Home traded productivity Permanent increase in Home traded productivity Temporary increase in Foreign patience
31
Figure 4: Temporary increase in Home traded productivity
θ = 1 θ = 3 θ = 6 θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0, β0 = 0.95
-0,25
-0,20
-0,15
-0,10
-0,05
0,00
0,05
1 2 3 4 5 6 7 8
lambda
Consumption
0,0
0,5
1,0
1,5
2,0
1 2 3 4 5 6 7 8
lambda
Savings
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Terms-of-trade
-0,35
-0,30
-0,25
-0,20
-0,15
-0,10
-0,05
0,001 2 3 4 5 6 7 8
lambda
Real exchange rate
0,0
0,1
0,2
0,3
0,4
0,5
0,6
1 2 3 4 5 6 7 8
lambda
Relative price of non-traded
-0,05
0,00
0,05
0,10
0,15
0,20
0,25
1 2 3 4 5 6 7 8
lambda
Real (cons) interest rate
32
Figure 4 (cont.): Temporary increase in Home traded productivity
θ = 1 θ = 3 θ = 6 θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0, β0 = 0.95
0,0
0,5
1,0
1,5
2,0
2,5
1 2 3 4 5 6 7 8
lambda
Traded output
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,0
0,1
1 2 3 4 5 6 7 8
lambda
Non-traded output
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Total output (traded units)
-0,9
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Labor in non-traded sector
33
Figure 5: Permanent increase in Home traded productivity
θ = 1 θ = 3 θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0, β0 = 0.95
0,00
0,05
0,10
0,15
0,20
0,25
0,30
0,35
1 2 3 4 5 6 7 8
lambda
Consumption
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Terms-of-trade
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Real exchange rate
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Relative price of non-traded
34
Figure 5 (cont.): Permanent increase in Home traded productivity
θ = 1 θ = 3 θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0, β0 = 0.95
0,0
0,5
1,0
1,5
2,0
2,5
1 2 3 4 5 6 7 8
lambda
Traded output
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Non-traded output
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Total output (traded units)
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Labor in non-traded sector
35
Figure 6: Temporary increase in Foreign patience
θ = 1 θ = 3 θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0, β0 = 0.95
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
1 2 3 4 5 6 7 8
lambda
Consumption
-3,0
-2,5
-2,0
-1,5
-1,0
-0,5
0,01 2 3 4 5 6 7 8
lambda
Savings
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Terms-of-trade
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Real exchange rate
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
1 2 3 4 5 6 7 8
lambda
Relative price of non-traded
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Real (cons) interest rate
36
Figure 6 (cont.): Temporary increase in Foreign patience
θ = 1 θ = 3 θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0, β0 = 0.95
-0,9
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Traded output
0,0
0,1
0,2
0,3
0,4
1 2 3 4 5 6 7 8
lambda
Non-traded output
0,0
0,1
0,2
0,3
0,4
1 2 3 4 5 6 7 8
lambda
Total output (traded units)
0,0
0,1
0,2
0,3
0,4
0,5
0,6
1 2 3 4 5 6 7 8
lambda
Labor in non-traded sector
37
Figure 7: Impact on short run Home variables All shocks imply a unit Home real appreciation in the short run
Consumption
Savings
θ = 3 θ = 6 θ = infinity
-0,8
-0,6
-0,4
-0,2
0,0
0,2
0,4
0,6
0,8
1 2 3 4 5 6 7 8
lambda
Temporary increase in A_H
0,0
0,5
1,0
1,5
2,0
1 2 3 4 5 6 7 8lambda
Permanent increase in A_H
0
1
2
3
4
1 2 3 4 5 6 7 8lambda
Temporary increase in beta_F
0
2
4
6
8
10
12
1 2 3 4 5 6 7 8
lambda
Temporary increase in A_H
-16
-14
-12
-10
-8
-6
-4
-2
01 2 3 4 5 6 7 8
lambda
Temporary increase in beta_F
38
Figure 7 (cont.): Impact on short run Home variables All shocks imply a unit Home real appreciation in the short run Terms-of-trade
Relative price of non-traded
θ = 3 θ = 6 θ = infinity
05
10152025303540
1 2 3 4 5 6 7 8
lambda
Temporary increase in A_H
02468
101214161820
1 2 3 4 5 6 7 8lambda
Permanent increase in A_H
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Temporary increase in beta_F
0,0
0,5
1,0
1,5
2,0
2,5
3,0
1 2 3 4 5 6 7 8
lambda
Temporary increase in A_H
0,0
0,5
1,0
1,5
2,0
2,5
1 2 3 4 5 6 7 8lambda
Permanent increase in A_H
0,0
0,5
1,0
1,5
2,0
1 2 3 4 5 6 7 8lambda
Temporary increase in beta_F
39
Figure 7 (cont.): Impact on short run Home variables All shocks imply a unit Home real appreciation in the short run
Non-traded output
Total output (traded units)
θ = 3 θ = 6 θ = infinity
-7
-6
-5
-4
-3
-2
-1
01 2 3 4 5 6 7 8
lambda
Temporary increase in A_H
-3,5
-3,0
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1 2 3 4 5 6 7 8
lambda
Permanent increase in A_H
0,00
0,25
0,50
0,75
1,00
1 2 3 4 5 6 7 8lambda
Temporary increase in beta_F
05
10152025303540
1 2 3 4 5 6 7 8
lambda
Temporary increase in A_H
0
5
10
15
20
25
1 2 3 4 5 6 7 8lambda
Permanent increase in A_H
0,00,10,20,30,40,50,60,70,80,91,0
1 2 3 4 5 6 7 8lambda
Temporary increase in beta_F
40
Figure 8: Temporary increase in Home traded productivity, with bias
θ = 1
θ = 3
θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0.3, β0 = 0.95
-0,25-0,20
-0,15-0,10-0,05
0,000,05
0,100,15
0,200,25
1 2 3 4 5 6 7 8
lambda
Consumption
-1,0
-0,5
0,0
0,5
1,0
1,5
2,0
1 2 3 4 5 6 7 8
lambda
Savings
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Terms-of-trade
-0,40
-0,30
-0,20
-0,10
0,00
0,10
0,20
0,30
1 2 3 4 5 6 7 8
lambda
Real exchange rate
0,0
0,1
0,2
0,3
0,4
0,5
0,6
1 2 3 4 5 6 7 8
lambda
Relative price of non-traded
-0,30
-0,20
-0,10
0,00
0,10
0,20
0,30
1 2 3 4 5 6 7 8
lambda
Real (cons) interest rate
41
Figure 8 (cont.): Temporary increase in Home traded productivity, with bias
θ = 1
θ = 3
θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0.3, β0 = 0.95
0,0
0,5
1,0
1,5
2,0
2,5
1 2 3 4 5 6 7 8
lambda
Traded output
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Non-traded output
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Total output (traded units)
-0,9
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Labor in non-traded sector
42
Figure 9: Permanent increase in Home traded productivity, with bias
θ = 1
θ = 3
θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0.3, β0 = 0.95
0,00
0,05
0,10
0,15
0,20
0,25
0,30
0,35
1 2 3 4 5 6 7 8
lambda
Consumption
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
1 2 3 4 5 6 7 8
lambda
Terms-of-trade
-0,8
-0,6
-0,4
-0,2
0,0
0,2
0,4
1 2 3 4 5 6 7 8
lambda
Real exchange rate
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Relative price of non-traded
43
Figure 9 (cont.): Permanent increase in Home traded productivity, with bias
θ = 1
θ = 3
θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0.3, β0 = 0.95
0,0
0,5
1,0
1,5
2,0
2,5
1 2 3 4 5 6 7 8
lambda
Traded output
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Non-traded output
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1 2 3 4 5 6 7 8
lambda
Total output (traded units)
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Labor in non-traded sector
44
Figure 10: Temporary increase in Foreign patience, with bias
θ = 1
θ = 3
θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0.3, β0 = 0.95
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
1 2 3 4 5 6 7 8
lambda
Consumption
-3,0
-2,5
-2,0
-1,5
-1,0
-0,5
0,01 2 3 4 5 6 7 8
lambda
Savings
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Terms-of-trade
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Real exchange rate
0,0
0,1
0,2
0,3
0,4
0,5
0,6
1 2 3 4 5 6 7 8
lambda
Relative price of non-traded
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0,01 2 3 4 5 6 7 8
lambda
Real (cons) interest rate
45
Figure 10 (cont.): Temporary increase in Foreign patience, with bias
θ = 1
θ = 3
θ = 6
θ = infinity Parameters: n = 0.1, γ = 0.3, α = 0.3, χ = 0.3, β0 = 0.95
-1,0
-0,8
-0,6
-0,4
-0,2
0,0
0,2
1 2 3 4 5 6 7 8
lambda
Traded output
-0,1
0,0
0,1
0,2
0,3
0,4
1 2 3 4 5 6 7 8
lambda
Non-traded output
-0,1
0,0
0,1
0,2
0,3
0,4
1 2 3 4 5 6 7 8
lambda
Total output (traded units)
-0,2
-0,1
0,0
0,1
0,2
0,3
0,4
0,5
0,6
1 2 3 4 5 6 7 8
lambda
Labor in non-traded sector
46
Appendix I: selected quotes on exchange rate appreciations
JAPAN Sayuri Shirai: The euro area crisis, the flight-to-safety premium, and cooperation and coordination
among central banks
Remarks by Ms Sayuri Shirai, Member of the Policy Board of the Bank of Japan, at the workshop co-
hosted by the Asian Development Bank Institute and the Reinventing Bretton Woods Committee,
“Adjusting the World to the New Realities of the International Financial System”, Tokyo, 12 October
2012.
“Consequently, capital inflows into securities investment from abroad have contributed to the
appreciation of the exchange rates of some of these countries, as typically seen in the case of Japan.
This has adversely affected the export manufacturing sector, which had been recovering from the
supply-chain disruption and resultant plunge in exports caused by the Great East Japan Earthquake on
March 11, 2011.”
Masaaki Shirakawa: Reviewing the economies of Switzerland and Japan Remarks by Mr Masaaki Shirakawa, Governor of the Bank of Japan, at the 30th Anniversary Luncheon of
the Swiss Chamber of Commerce and Industry in Japan (SCCIJ), Tokyo, 10 October 2012.
“More recently, between December 2007 and August 2011, the real effective exchange rates for
Switzerland and Japan moved in tandem with each other, appreciating by 16.9 percent and 23.3 percent,
respectively. These movements of the last few years are to an extent representative of the fallouts from the
Great Financial Crisis. The prevalent mood of risk aversion resulted in the unwinding of carry trades.
Both countries are now also confronted by the so-called flight-to-safety capital flows. The appreciation
of the two currencies coincided with the most severe financial and economic dislocations in the
advanced economies.”
Ryuzo Miyao: Economic activity, prices, and monetary policy Speech by Mr Ryuzo Miyao, Member of the Policy Board of the Bank of Japan, at a meeting with
business leaders, Yamaguchi, 5 September 2012.
“Second, there is a risk of further appreciation of the yen. While a positive case can be made for yen
appreciation to a certain extent, excessive appreciation will worsen the competitiveness and profits of
exporting firms again and act as a headwind against firms. If a trend of falling stock prices intensifies,
together with the trend of yen appreciation, then firms’ and households’ confidence will deteriorate
and currently solid business fixed investment plans and private consumption will be contained, which will weigh on Japan’s economic recovery.”
Yoshihisa Morimoto: Economic activity and prices in Japan and monetary policy Speech by Mr Yoshihisa Morimoto, Member of the Policy Board of the Bank of Japan, at a meeting with
business leaders, Ishikawa, 2 August 2012.
“Thereafter, however, it (the economy) remained more or less flat on the whole until around the early
springof 2012 mainly due to the adverse effects of the slowdown in overseas economies and the
appreciation of the yen.”
Hirohide Yamaguchi: European debt problem and its impact on Asia
47
Speech by Mr Hirohide Yamaguchi, Deputy Governor of the Bank of Japan, at the 8th Tokyo-Beijing
Forum, Tokyo, 2 July 2012.
“Changes in foreign exchange rates, for example, a substantial yen appreciation might, especially for
Japan’s economy, worsen corporate profits and put downward pressure on economic activity.”
Masaaki Shirakawa: Japan’s economy and monetary policy Speech by Mr Masaaki Shirakawa, Governor of the Bank of Japan, at a meeting held by the
Naigai Josei Chousa Kai (Research Institute of Japan), Tokyo, 4 June 2012.
“After the beginning of autumn 2011, however, economic activity temporarily became flat due to the
effects of the slowdown in overseas economies, the appreciation of the yen, and the flooding in
Thailand.”
Sayuri Shirai: Recent global economic developments and monetary policy in Japan – strengthening
Japan’s growth momentum through opportunities in emerging Asia Speech by Ms Sayuri Shirai, Member of the Policy Board of the Bank of Japan, at a meeting with
business leaders, Akita, 10 May 2012.
“From the second half of fiscal 2011, economic activity remained more or less flat mainly due to an
adverse effect on exports and production of the slowdown in overseas economies and the appreciation
of the yen.”
Other countries
Emmanuel Tumusiime-Mutebile: Oil revenue management Remarks by Mr Emmanuel Tumusiime-Mutebile, Governor of the Bank of Uganda, at the Oil Revenue
Management Seminar, Kampala, 27 February 2012.
“The second issue which I want to highlight pertains to the potential for Dutch disease. The spending of
oil revenues is likely to cause a real exchange rate appreciation, which could damage the
competitiveness of the non oil traded goods sectors of the economy, including both exporters and
manufacturing firms which compete with imports.”
Mark Carney: Dutch disease Remarks by Mr Mark Carney, Governor of the Bank of Canada and Chairman of the Financial Stability
Board, to the Spruce Meadows Round Table, Calgary, Alberta, 7 September 2012
“Some regard Canada’s wealth of natural resources as a blessing. Others see it as a curse. The latter
look at the global commodity boom and make the grim diagnosis for Canada of “Dutch Disease.”1 They
dismiss the enormous benefits, including higher incomes and greater economic security, our bountiful
natural resources can provide. Their argument goes as follows: record-high commodity prices have led to
an appreciation of Canada’s exchange rate, which, in turn, is crowding out trade-sensitive sectors, particularly manufacturing. The disease is the notion that an ephemeral boom in one sector causes
permanent losses in others, in a dynamic that is net harmful for the Canadian economy.”
Jean-Pierre Danthine: Monetary policy is not almighty
Speech by Mr Jean-Pierre Danthine, Vice Chairman of the Governing Board of the Swiss National Bank,
at the Journée Solutions Bancaires, Geneva, 31 May 2012.
48
“An appreciation of the Swiss franc would again expose the Swiss economy to considerable risks and, once more, endanger both price stability and economic situation.”
Thomas Jordan: The Swiss economy and global economic outlook Introductory remarks by Mr Thomas Jordan, Chairman of the Governing Board of the Swiss National
Bank, at the Media News Conference of the Swiss National Bank, Berne,
14 June 2012.
“The Swiss National Bank (SNB) will maintain the minimum exchange rate of CHF 1.20 per euro and
will enforce it with the utmost determination. It remains prepared to buy foreign currency in unlimited
quantities for this purpose. Even at the current rate, the Swiss franc is still high. Another appreciation
would have a serious impact on both prices and the economy in Switzerland. The SNB will not tolerate
this. If necessary, it stands ready to take further measures at any time.”
Christian Noyer: Monetary policy, the French economy and its outlook
Introductory letter by Mr Christian Noyer, Governor of the Bank of France, to the Bank of France’s
Annual Report 2006, submitted to the President of the French Republic, the President of the Senate, and
the President of the National Assembly, 1 August 2007.
“It is clear that the price-competitiveness of French industries has deteriorated significantly in recent
years. Has the euro’s appreciation played a role in this? On the one hand, it undoubtedly penalises
export sectors whose competitors are located in other monetary areas. But, on the other hand, it
benefits those sectors which are large consumers of imported commodities. At this stage, the overall effect on France’s growth and external balance is not clearly apparent. Furthermore, two observations
must be made: first, it is mainly intra-euro area trade that is worsening; second, many of France’s
European partners faced with the same constraints are performing better”
Axel A Weber: Financial markets, economic forecast and monetary policy Speech by Professor Axel A Weber, President of the Deutsche Bundesbank, at the British Chamber of
Commerce in Germany, Frankfurt am Main, 17 April 2008.
“Third, the euro’s appreciation should not only be viewed from the perspective of external demand. A
strong euro simultaneously dampens the prices of import goods. This helps to curb the inflationary pressures resulting from strong rises in energy and food prices and thus has a stabilising effect”
Frederic S Mishkin: Globalization, macroeconomic performance, and monetary policy
Member of the Board of Governors of the US Federal Reserve System, at the Domestic Prices in an
Integrated World Economy Conference, Board of Governors of the Federal Reserve System, Washington
DC, 27 September 2007.
“An appreciation of the dollar, in turn, restrains exports (because the price of U.S. goods rises when
measured in foreign currencies) and stimulates imports (because imports become cheaper in dollar terms). The resulting decrease in net exports implies a reduction in aggregate demand.”
49
Appendix II: complete solution of the model
A II.1 Consumption allocation
The allocation of consumption in the Home country is:
( )[ ] [ ][ ] [ ]
[ ] tttNtN
tttTtTtFtF
tttTtTtHtH
CPPC
CPPPPnC
CPPPPnnC
λ
λθ
λθ
γ
χγ
χγ
−
−−
−−
−=
−−=
−+=
/)1(
//)1)(1(
//)1(
,,
,,,,
,,,,
The corresponding relations in the Foreign country are:
[ ] [ ]( )[ ] [ ]
[ ] ***
,
*
,
***
,
*
,,
*
,
***
,
*
,,
*
,
/)1(
//1
//)1(
tttNtN
tttTtTtFtF
tttTtTtHtH
CPPC
CPPPPnnC
CPPPPnC
λ
λθ
λθ
γ
χγ
χγ
−
−−
−−
−=
+−=
−=
A II.2 Market clearing and intertemporal constraints
Using the intratemporal consumption allocation, the clearing conditions for the non-traded
goods (1) are written as:
( )( )( )
( )( )( )
αλλ
λ
λ
αλλ
λ
λ
γγγ
γγγ
−−
−
−
−−
−
−
=
−+−−
=
−+−
1*
,
*
,
*
1
1*
1*
1
,,
1
1
1
)(1
)1)(1(
)(1
)1(
tNtNt
t
t
tNtNt
t
t
LACR
Rn
LACR
Rn
The Home intertemporal constraint in terms of traded goods (2) is written as:
( ) ( )[ ]( )( )[ ]
1
1
,,
11
1
11
)1()(
1
)1)(1()1(+
−
−−
−−
−++−=−+
−−+−+ttttttNtHt
t
t BnTBTrnLnAC
R
Tnnnn
α
λλ
λ
θθ
γγ
χχγ
The Foreign intertemporal constraint in terms of traded goods (3) is written as:
( )( )[ ]( )( )[ ] 1
1*
,
*
,
*
11*
1
11
)1()(
1
1)1()1( +
−
−−
−−
++−−=−+
+−+−− ttttttNtFtt
t
t BnTBTrnLnATC
R
Tnnnn
α
λλ
λ
θθ
γγ
χχγ
The clearing of the market for Home traded goods is written as:
50
( ) ( )[ ] ( )( )[ ]
( )( )[ ] ( )( )[ ]*
11*11
1111
1
,,
1
1
1)1(
)1()1(
1
1
)1)(1()1(
)1()(
t
tt
t
tt
tNtH
C
RTnnn
nn
C
RTnnn
nnnLnA
−−−−
−−−−
−
−++−+−
−−+
−+−−+−+
−+=−
λλ
λθθ
θ
λλ
λθθ
θ
α
γγχχ
χγ
γγχχ
χγ
A II.3 Real interest rate and labor allocation
Using the intratemporal consumption allocation, the real interest rates in terms of
consumption baskets (6) are written as:
( )( )( )( )
( ) ( )( ) ( )
( )( )( )( )
( )( )( )( )
θ
θ
θλ
λ
λ
θ
θ
θλ
λ
λ
χχχχ
γγγγ
χχχχ
γγγγ
−
−+
−−
−+
−+
++
−
−+
−−
−+
−+
++
+−+−+−+−
−+
−++=+
−−+−+−−+−+
−+−+
+=+
1
1
1
1
11
1
1*
1
1*
1
1
*
1
1
1
1
1
11
1
1
1
1
1
11
1)1(
1)1(
1
1)1(1
)1)(1()1(
)1)(1()1(
1
1)1(1
t
t
t
t
t
t
t
C
t
t
t
t
t
t
t
t
C
t
Tnnn
Tnnn
R
R
T
Trr
Tnnn
Tnnn
R
R
T
Trr
The allocations of labor (7)-(8) are written as:
( ) ( )[ ]( )( )[ ] αθθα
αθθα
χχ
χχ
−−−−
−−−−
+−+−=−−
−−+−+=−
)(1
1)1()1(
)()1)(1()1()(
*
,
*
,
*1
11*
,
*
,
,,1
11
,,
tNtNt
t
ttNtF
tNtNtttNtH
LART
TnnnLnA
LARTnnnLnA
A II.4 Log-linear approximations
In terms of log-linear approximations, we write the market clearing conditions for non-
traded goods (1) as:
*
,
*
,
**
,,ˆ)1(ˆˆˆ ; ˆ)1(ˆˆˆ
tNtNtttNtNtt LACRLACR αλγαλγ −+=+−−+=+−
The intertemporal constraints (2)-(3) are:
1
0
*
,
*
,
**
1
0
,,
ˆ1
ˆ1
1ˆ1
)1(ˆˆ)1(ˆ)1(ˆ
ˆˆ1ˆ1)1(ˆˆ)1(ˆ)1)(1(ˆ
+
+
−+
−−−−−=−+−−
−+−−−=−+−−+
tttNtFttt
tttNtHttt
Bn
nB
n
nLARTnC
BBLARTnC
βγγαγλχ
βγγαγλχ
In the short run we have sstt BBB ˆˆ ,0ˆ
1 == + , while in the long run sstt BBB ˆˆˆ
1 == + . The market
clearing for the Home traded good (4) is:
51
[ ]*
*
2
,,
ˆ)1)(1(ˆ))1((
ˆ)1)(1(ˆ))1(()1(
ˆ)1)(1(ˆ1)1(ˆ
tt
tt
ttNtH
CnCnn
RnRnn
TnLA
χχ
χχγλ
χθγ
γα
−−+−++
−−+−+−+
−−=−−−
The Euler equations (5) are (where 0011 /)1( ββ−−= ++
C
t
C
t rdr ):
*
101,
**
1101,1ˆˆˆ ; ˆˆˆ C
ttFtt
C
ttHtt drCCdrCC ++++++ ++=++= ββββ
In the long run we have 0ˆ ,ˆˆˆ11,1 ==== +++
C
ttHsstt drCCC β and 0ˆ ,ˆˆˆ *
11,
***
1 ==== +++C
ttFsstt drCCC β .
The real interest rates in terms of consumption baskets (6) are:
)ˆˆ)(1()ˆˆ)(1(
)ˆˆ)(1()ˆˆ)()1((
**
1110
*
10
111010
ttttt
C
t
ttttt
C
t
RRTTndrdr
RRTTnndrdr
−−−−−+=
−−−−−++=
++++
++++
γχββγχββ
The labor allocations (7)-(8) are:
TnARLATnARLA tNttNtFttNttNtHˆ)1(ˆˆˆ1ˆ ; ˆ)1)(1(ˆˆˆ1ˆ *
,
**
,
*
,,,, χγ
αχγ
α −−+=+−−++=+
A II.5 A simple case
We set ∞→θ and λ = 1. The approximated non-traded market clearing conditions (1)
are then:
*
,
*
,
**
,,ˆ)1(ˆˆˆ ; ˆ)1(ˆˆˆ
tNtNtttNtNtt LACRLACR αγαγ −+=+−−+=+−
The intertemporal constraints (2)-(3) are:
1
0
*
,
*
,
**
1
0
,,
ˆ1
ˆ1
1ˆ1
)1(ˆˆ)1(ˆ
ˆˆ1ˆ1)1(ˆˆ)1(ˆ
+
+
−+
−−−−−=−+
−+−−−=−+
tttNtFtt
tttNtHtt
Bn
nB
n
nLARC
BBLARC
βγγαγ
βγγαγ
The market clearing for the Home traded good (4) becomes redundant. The Euler equations (5)
are as in the general case, and the real interest rates in terms of consumption baskets (6) are:
)ˆˆ)(1( ; )ˆˆ)(1( **
110
*
1011010 ttt
C
tttt
C
t RRdrdrRRdrdr −−−=−−−= ++++++ γββγββ
The labor allocations (7)-(8) are:
*
,
**
,
*
,,,,ˆˆˆ1ˆ ; ˆˆˆ1ˆ
tNttNtFtNttNtH ARLAARLA +=++=+γ
αγ
α
In the long-run the economy reaches a new steady-state, with the cross-country asset
holdings ssB as a state variable. Consumption is higher in the Home country, relative to the
52
Foreign one, if the Home country has higher productivity of is a net creditor ( 0ˆ >ssB ). A net
creditor position or an advantage in traded sector productivity raises the price of non-traded
goods in the Home country, which translates into an appreciated currency. A net creditor position
also shifts labor towards the non-traded sector:
n
BLL
RRQ
n
BAAAARR
n
BAAAACC
ss
ssNssN
ssssss
ss
ssNssNssFssHssss
ss
ssNssNssFssHssss
−−
=−
−−−=
−−
+−−−=−
−−
+−−+−=−
1
ˆ1ˆˆ
)ˆˆ)(1(ˆ
1
ˆ1)ˆˆ()ˆˆ(ˆˆ
1
ˆ1)ˆˆ)(1()ˆˆ(ˆˆ
0
0*
,,
*
0
0*
,,
*
,,
*
0
0*
,,
*
,,
*
ββγ
γ
ββα
ββγγγ
We denote worldwide averages with a w superscript, with for instance
*
,,,ˆ)1(ˆˆ
ssFssH
w
ssT AnAnA −+= . The solution is then: w
ssN
w
ssT
w
ss AAC ,,ˆ)1(ˆˆ γγ −+= , w
ssN
w
ssT
w
ss AAR ,,ˆˆˆ −= and
0ˆ, =w
ssNL .
The short-run cross-country solution is:
−+−−
−−−−−
−+−−=−
−+−=−
−−−=
−+−−−−=−
−+−−−+−=−
−+=
−
)ˆˆ()ˆˆ(
)ˆˆ()ˆˆ()1(ˆ
)1(
)1()(
ˆ)1(
ˆˆ
)ˆˆ)(1(ˆ
ˆ)1(
)ˆˆ()ˆˆ(ˆˆ
ˆ)1(
)ˆˆ)(1()ˆˆ(ˆˆ
ˆ)1(1
ˆ
**
,,
**
,,*
0
0*
,,
*
0***
0***
0
NNssNssN
FHssFssHCC
ssNssN
ss
NNFH
NNFH
ss
AAAA
AAAAZdrdr
ZLL
RRQ
ZAAAARR
ZAAAACC
Zn
B
γαγγ
αγβ
αγγβγ
γαγγ
βα
αγγβγγγ
αγγβ
where )ˆˆ()ˆˆ()ˆˆ(ˆ **
,, FHssFssHFH AAAAZ −+−−−= ββ , so 0ˆ >Z indicates that the Home country is
relatively more patient (FH ββ ˆˆ > ), or that it has a temporary productivity advantage in the traded
sector ( *
,,
* ˆˆˆˆssFssHFH AAAA −>− ).
53
In worldwide terms, we have w
N
w
T
wAAC ˆ)1(ˆˆ γγ −+= , w
Ns
w
T
wAAR ˆˆˆ −= and 0ˆ =w
NL . In
addition, the real interest rates are higher when agents become more impatient on average or
future productivity levels exceed current ones:
)ˆˆ)(1()ˆˆ(ˆˆ
)ˆˆ(ˆˆ
,,0
,0
w
N
w
ssN
w
T
w
ssT
wwCw
w
T
w
ssT
ww
AAAACdr
AACdr
−−+−+−==
−+−==
γγββ
ββ
54
Table A1: list of country episodes selected with the criteria of lending booms only and lending booms and strong appreciation combined
Country Years identified according to Country Years identified according to
LB LB and ABC LB LB and ABC
Australia 2003 - 2011 2004 - 2004 Pakistan 1997 - 2010 2004 - 2004
Austria 1994 - 2011 Peru 1997 - 1997
Brazil 1989 - 1989 1999 - 2011
1993 - 1993 Saudi Arabia 1999 - 2011
Chile 1998 - 2011 South Africa 1992 - 2010
Colombia 1997 - 1999 2004 - 2006 Spain 1989 - 1994
2001 - 2008 2002 - 2011
Denmark 2000 - 2011 Sweden 1998 - 1999
Egypt 2001 - 2011 2003 - 2004 2001 - 2003
Euro Area 2001 - 2011 Sweden 2008 - 2011
Germany 1978 - 1984 Switzerland 1980 - 2011
2006 - 2011 Thailand 1992 - 2000 2007 - 2007
Greece 2010 - 2011 2002 - 2005
Hong Kong 1990 - 2011 2007 - 2011
Ireland 1999 - 2011 2003 - 2004 United Kingdom 1989 - 2011 1997 - 1998
Israel 2008 - 2008 2008 - 2008 United States 1972 - 1973 1983 - 1983
Italy 2007 - 2011 United States 1983 - 2011 1997 - 1998
Japan 1970 - 2011 1984 - 1987
1992 - 1994
Korea 2007 - 2011
Netherlands 1986 - 1986
1992 - 2011
New Zealand 1996 - 1997
1999 - 2011
55
Table A2: list of country episodes selected with the criteria of capital flow bonanzas only and capital flow bonanzas and strong appreciation
combined
Country Years identified according to Country Years identified according to
CFB CFB and ABC CFB CFB and ABC
Australia 1971 - 1971 New Zealand 1989 - 1989
1987 - 1987 1996 - 1996
1998 - 1998 1998 - 2002
2001 - 2003 2005 - 2005
2006 - 2009 2009 - 2009
Austria 2000 - 2001 Norway 2000 - 2000
2007 - 2007 2005 - 2007
Belgium 2002 - 2010 2009 - 2009
Brazil 1999 - 2001 2007 - 2007 Pakistan 1984 - 1985
2007 - 2007 1990 - 1990
Canada 1997 - 1997 2008 - 2008 2000 - 2000
2005 - 2008 Peru 2000 - 2000
Chile 1993 - 1998 2005 - 2007
2005 - 2008 Philippines 1994 - 1996
China 1987 - 1987 1994 - 1994 2005 - 2005
1989 - 1989 2007 - 2010
1994 - 1994 Poland 1996 - 1996 2007 - 2007
1996 - 2001 2006 - 2007
2003 - 2010 Portugal 2006 - 2007
Colombia 1998 - 2001 2006 - 2006 Russia 1998 - 2000
Colombia 2006 - 2008 2004 - 2004
Costa Rica 1998 - 1998 2006 - 2007
2004 - 2009 Saudi Arabia 1997 - 1998
Czech Rep. 1993 - 1993 2002 - 2002 2000 - 2001
1995 - 1995 2003 - 2006
1998 - 2002 2008 - 2008
2004 - 2005 Singapore 2007 - 2007
2007 - 2007 South Africa 1997 - 1999
56
Table A2(cont.): list of country episodes selected with the criteria of capital flow bonanzas only and capital flow bonanzas and strong appreciation
combined
Country Years identified according to Country Years identified according to
CFB CFB and ABC CFB CFB and ABC
Czech Rep. 1993 - 1993 2002 - 2002 2000 - 2001
1995 - 1995 2003 - 2006
1998 - 2002 2008 - 2008
2004 - 2005 Singapore 2007 - 2007
2007 - 2007 South Africa 1997 - 1999
Denmark 1999 - 2001 2001 - 2001
2007 - 2007 2005 - 2007
Egypt 1998 - 2007 2003 - 2004 2009 - 2009
2009 - 2010 Spain 1994 - 1995
Euro Area 1998 - 2002 1998 - 2000
2007 - 2008 2002 - 2002
Finland 1979 - 1979 2006 - 2008
2005 - 2008 Sri Lanka 1982 - 1982
France 1997 - 2002 1984 - 1984
2005 - 2007 Sweden 1972 - 2007 1982 - 1982
Germany 1999 - 2001 2009 - 2010 1990 - 1991
2005 - 2006 Switzerland 1985 - 1986
2009 - 2009 1989 - 1989
Greece 2006 - 2007 1996 - 2002
Guatemala 1988 - 1988 2006 - 2009
Hong Kong 1998 - 2010 Thailand 1989 - 1989 2007 - 2007
Hungary 1991 - 1991 1997 - 1999
1993 - 1993 2001 - 2001
1995 - 2002 2003 - 2007
2004 - 2009 Tunisia 2006 - 2007
Indonesia 2007 - 2007 1981 - 1982
2009 - 2009 2006 - 2006
2008 - 2008
57
Table A2(cont.): list of country episodes selected with the criteria of capital flow bonanzas only and capital flow bonanzas and strong appreciation
combined
Country Years identified according to Country Years identified according to
CFB CFB and ABC CFB CFB and ABC
Ireland 1998 - 2007 2003 - 2004 United Kingdom 1987 - 1987 1998 - 1998
2009 - 2010 1989 - 1989
Israel 1999 - 2000 2008 - 2008 1993 - 1993
2004 - 2008 1998 - 2001
Korea 1999 - 2000 2005 - 2009
Mexico 2003 - 2003 United States 1999 - 2000
Morocco 2001 - 2001 Uruguay 2005 - 2010
Netherlands 1974 - 1974
1981 - 1983
1989 - 1997
2004 - 2004
2006 - 2007
top related